325 1 References and Notes RH Atalla,* JC Gast DW Sindorf, VJ

partial support of the work from institutional funds of The. Institute of Paper Chemistry and the Colorado State Univer- sity Experiment Station. Drs...
1 downloads 0 Views 298KB Size
325 1

Communications to the Editor

Acknowledgments, The authors gratefully acknowledge partial support of the work from institutional funds of The Institute of Paper Chemistry and the Colorado State University Experiment Station. Drs. W. L. Earl and D. L. VanderHart kindly provided a preprint of the accompanying communication. References and Notes (1) Jones, D. W. In "Cellulose and CelluloseDerivatives", Bikales, N. M., Segal. L., Eds.: Wiley-lnterscience, New York, 1971; Part IV, p 117. (2) Ellefsen, 0.; Tonnesen, B. A. Reference 1, p 151. (3) Tonnesen, B. A.; Ellefsen, 0. Reference 1, p 265. (4) Atalla, R. H.; Dimick, B. E. Carbohydr. Res. 1975, 39, C1. (5) Atalla, R. H. Proc. Cellulose Conf., 8th, 1976, Appl. Polym. Symp. 1978, No. 28, 659. (6) Atalla. R. H ACSAdv. Chem. Ser. 1979, No. 181, 55. (7) Elsevier: New . . Kakudo, M.: Nobutami. K. "X-ray Diffractlon by. Polymers"; . York, 1972; p 285. (8) Petitpas, T.; Oberlin, M.; Mering, J. J. Polymer Sci. S 1983, 2, 423. (9) Norman, M. Text. Res. J. 1963, 33, 711. (10) Gardner, K. H.; Blackwell, J. Biopolymers 1974, 13, 1975. (1 1) Sarko. A,: Muaoli. R. h4acromolecules 1974. 7. 486. i12j S+aefer,'J.; Srejskael, E. 0. Top. Carbon-i3 hMR Spectrosc. 1979, 3. (13) Miknls, F. P.; Bartuska, V. J.: Maciel, G. E. Am. Lab., in press. (14) Maciel, G.E.; Bartuska, V. J.; Mlknis, F. P. Fuel, in press. (15) Atalla, R. H.; Nagel, S. C. Science 1974, 785, 522. (16) Atalla, R. H.; Dimick, 8.E.; Nagel, S.C. ACS Symp. Ser. 1977, No. 48, 30. (17) Gast. J. C.: Atalla, R. H.; McKelvey. R. D. Carbohydr. Res., In press.

R. H. Atalla,* J. C. Gast Institute of Paper Chemistry Appleton, Wisconsin 5491 2 D. W. Sindorf, V. J. Bartuska, C. E. Maciel* Department of Chemistry,'Colorado State University Fort Collins, Colorado 80523 Received September 28, 1979

High Resolution, Magic Angle Sample Spinning I3C NMR of Solid Cellulose I' Sir: Recent developments in I3C N M R in solids have made it possible to obtain spectra with lines narrow enough to distinguish resonances due to individual carbons in polymers.2 W e have combined the techniques of cross polarization, high power d e ~ o u p l i n gand , ~ magic angle sample spinning4 to study a variety of solid samples, This report is a preliminary account of studies aimed a t characterizing several preparations of cellulose. W e have measured the I3C N M R spectra of glucose and cellobiose in the solid state and compared the spectra to peak assignments of the solution N M R spectra of the same comp o u n d ~ By . ~ analogy, it is relatively easy to identify the peaks in the solid-state spectrum of cellulose which correspond to carbons 1, 4, and 6. Figure 1 shows the structure of cellulose and the cross polarization/magic angle sample spinning (CP/MASS) spectrum of a sample of microcrystalline cellulose with the assigned peaks marked. Thi's spectrum is virtually identical with other spectra obtained from native celluloses in our laboratory. This spectrum is also the same as the spectrum of Whatman CF-1 reported by Atalla et a1.I It is significant that the peaks in the I3C CP MASS spectra of glucose and of cellobiose do not have exactly the same chemical shifts relative to Me4Si as the solution spectra of the same compounds. The differences in shifts might be attributed to packing effects in the solid or, more likely, to hydrogen bonding in the solid which is liable to be quite different from that in solution. The chemical similarity of carbons 2 , 3 , and 5 make it reasonable that their resonance peaks should overlap, but, as yet, it has not been possible to make further assignments in this region. Integration of the peaks in Figure 1 gives a ratio of

C1

l

l

120 l l

l1001

180 1

60 1 1

40 i 1ppm(TMS) i I 1

1

Figure 1. The CP/MASS spectrum of a dried sample of microcrystalline cellulose 1. The chemical formula for cellulose is shown above. The , shown at the bottom of the figure with transverse relaxation times, T ~ care the regions of the spectrum to which they correspond. This spectrum i s the result of 7000 scans with 1024 data points per scan at a dwell time of 50 ps per point. The external applied field was 1.4 T . The I3C and ' H radio frequency fields were matched at 57 kHz. The time between successive scans was 3.6 s a n d the spinning frequency was 2100 Hz.

1:1:3:1 but, to obtain this ratio, the C-4 resonance was assumed to include both the broad peak located between 8 1 and 88 ppm and the sharper peak centered a t 90 ppm; similarly, the C-6 resonance included the high-field tail. Two explanations for the broad resonance centered a t 85 ppm and the upfield shoulder on C-6 were investigated. The first was that hydration of the cellulose resulted in shifts in C-4 and C-6 resonances due to hydrogen bonding with water molecules. A sample of microcrystalline cellulose6 was carefully dried by heating to 140 OC for -40 h under vacuum, followed by careful loading into a rotor which was sealed to prevent atmospheric water from hydrating the sample. The C P M A S S spectrum of the dried sample was indistinguishable from that of a hydrated sample. Thus we have concluded that the spectral features are not due to hydration of the cellulose. The second explanation for the broad resonances is an analogy to polyethylene where we have observed a single sharp resonance with a broad shoulder which could be attributed to noncrystalline carbons.' Relaxation time measurements have been used to establish that the noncrystalline regions of polyethylene exhibit much greater mobility than the crystalline regions.* We have investigated both the longitudinal, T I C and , transverse, T ~ crelaxation , times for the peaks observed in the 13C spectrum of the carefully dried sample of cellulose I. Saturation recovery experiments were employed to get a qualitative idea of the longitudinal relaxation times, TIC.^ The long Ti values observed in solid I3CN M R in general cause T I C measurements to suffer from poor signal to noise. As a result it was not possible to obtain quantitative values for the T I C values, but several qualitative observations were made. The overall relaxation time for recovery of longitudinal magnetization is tens of seconds. Within the accuracy of the data, all peaks in the observed spectrum relax with the same time constant with the exception of the upfield shoulder on (2-6. This shoulder has a somewhat shorter value of T I C indicating , increased molecular mobility.

This article not subject to U.S. Copyright. Published 1980 by the American Chemical Society

3252

Journal of the American Chemical Society

We have measured the transverse relaxation time, T ~ c for , the major features in the spectrum of the same dried sample of microcrystalline cellulose. This measurement involves a modification7 of the original echo sequence discussed by Hahn.’” Figure 1 includes the values obtained for the Tzc values, reported as effective full widths at half-maximum, Av(T2c) = (7rTzc)-’. The estimated errors in Au(T2c) are f 2 Hz. Most of these relaxation data were nominally exponential; the decay for C-6 was clearly nonexponential and could be decomposed into two components having associated line widths of 14 and 4 Hz. The faster relaxing component was identified to be the upfield tail of the resonance line. These relaxation studies permit several qualitative conclusions regarding the structure of cellulose I. The T I Cvalues observed and presumably the molecular motions are in a range between those observed in glassy synthetic polymers and highly crystalline materials.’ W e cannot exclude the possibility that the relaxation is due to motion of water molecules that remain after the drying process or motion of labile protons. Both the T I Cand T ~ data c show a faster relaxation for the carbons in the upfield tail of C-6 relative to the sharp peak. This suggests a greater mobility or more access to labile protons for those carbons in the tail as opposed to the sharp peak of C-6. The corresponding chemical shift may arise from differences in conformation, hydrogen bonding, or crystal packing. The observed full widths a t half height, Au*, in the C P MASS spectrum of cellulose I range from 30 to 85 Hz. This can be constrasted with measured Au(T2c) values of